LINGUIST List 23.1581

Thu Mar 29 2012

Review: Linguistic Theories; Syntax: Auer and Pfänder (2011)

Editor for this issue: Anja Wanner <anjalinguistlist.org>



Date: 29-Mar-2012
From: Thomas Hoffmann <thomas.hoffmannuni-osnabrueck.de>
Subject: Constructions: Emerging and Emergent
E-mail this message to a friend

Discuss this message

Announced at http://linguistlist.org/issues/22/22-3408.html

Editors: Auer, Peter and Pfänder, StefanTitle: Constructions: Emerging and EmergentSeries Title: De Gruyter linguae & litterae / Publications of the School ofLanguage and Literature Freiburg Institute for Advanced Studies 6Publisher: De Gruyter MoutonYEAR: 2011

Thomas Hoffmann, University of Osnabrück

SUMMARY

Ever since Chomsky's ''Aspects of the Theory of Syntax'' (1965), thecompetence-performance dichotomy has played a central role in syntacticresearch: On the one hand, mainstream generative grammarians claim thatperformance is influenced by "memory limitations, distractions, shifts ofattention and interest and errors (random and characteristic)" (Chomsky 1965: 3)and they therefore focus only on competence (mainly by eliciting introspectiondata). On the other hand, several researchers have pointed out that performanceis in fact much more systematic and rule-governed than mainstream generativegrammarians assume (Aarts 1991; Sampson 2001). Indeed, as the papers of thepresent volume argue, there are a great number of syntactic phenomena that canonly be adequately described and explained once their real, on-line use inactual discourse is taken into account.

''Constructions: Emerging and Emergent'' is a collection of papers that explicitlylook at syntactic phenomena in interaction. All of the authors adopt an approachthat treats language as a dynamic system and an adaptive resource ininteraction. Moreover, central to the volume are Hopper's notion of grammar asan emergent system as well as form-meaning pairings, i.e. constructions, aselementary building blocks of grammar.

In the first contribution, ''Constructions: Emergent or emerging'' (pp. 1-21), theeditors, Peter Auer and Stefan Pfänder, provide an introduction to the centraltopics and themes of the volume. In particular, they give an overview of theconcept of emergence outside the field of linguistics as well as Hopper's notionof emergent grammar. As they point out, emergent grammar for Hopper (e.g. 1987,1998, 2004) is not a stable system that exists outside of language use. Instead,speakers' previous experiences are claimed to be constantly updated in andadapted to specific interactional encounters. Closely related are approachesthat focus on the results of these online adaptations and on how thesediachronically affect the speakers' mental representations (an approach whichHopper calls 'emerging grammar'). However, Auer and Pfänder argue that thedifferences between emergent and emerging grammar approaches might not be asgreat as suggested by Hopper. In particular, they stress that emergingstructures are sedimented fragments that have been entrenched through emergentinteraction. Auer and Pfänder outline the role of 'constructions,' i.e. pairingsof form and meaning, which they treat as emergent gestalts (prefabricatedsyntactic templates that arise through interaction). Finally, Auer and Pfändergive an overview of the various chapters of the volume.

The first of these contributions is Paul Hopper's 'Emergent grammar andtemporality in interactional linguistics' (pp. 22-44). Hopper emphasises howtemporality in interaction crucially shapes constructions. In particular, helooks at the ''such a/an'' construction (''Sam … has taken such an interest in thisretirement bit … That it- it really surprises me'', p.31) and sluicing (''Theyargued but I don't know what about.'') in emergent discourse. He explores'structuration' (i.e. emergence) of constructions in context, particularly thecreative and interactional meaningful blending of fixed formulae (which allowthe hearer to project potentially following constructions at any point in time).

Simona Pekarek Doehler turns to left and right dislocation constructions inFrench in her article 'Emergent grammar for all practical purposes' (pp. 45-87).She shows the inadequacy of movement analyses, highlighting instead thedifferent discourse management functions of these two structures in on-linesyntax: left-dislocated elements can be used as try markers, i.e. attempts toelicit a response indicating recognition from the hearer before the speakercontinues (A: ''l'acqua B: ouais A: …on sait que c'est de l'eau'' 'A: water B:yeah A: …one knows that it's water'; p. 59). Right-dislocatedelements, on the other hand, reemphasise a turn-relevant point that the hearerdid not pick up on (A: ''et tu le détestes'' (longer pause) … ''euh: ton: parentqui a- … dé vié'' 'and you hate him (longer pause) … your parent who has gottenon the wrong track'; p. 70). Thus, right-dislocated elements can be added toutterances if a speaker wants to indicate that he or she requires feedback fromthe hearer. Consequently, it is the recurrent interactional needs that lead tothe entrenchment of these constructional schemata.

Turn-taking also plays an important role in Arnulf Deppermann's 'Constructionsvs. lexical items as sources of complex meanings' (88-126). Deppermanninvestigates the interplay of lexical meaning potential and on-going adaptationswithin real-time conversations, focusing on two German constructions with''verstehen'' 'know/understand' (''Verstehst du?'' 'Do you understand?' and ''(NounPhrase) nicht verstehen (können) (Complement Clause)'' 'can/do not understand(Noun Phrase/Complement Clause)'. He stresses that the instantiations of bothconstructions have a wide range of discourse functions (including signallingconcern of insufficient hearer uptake, problems of formulation or refocusing inthe case of ''Verstehst du?'' or pre-disagreement and reproach in the ''(NounPhrase) nicht verstehen (können) (Complement Clause)'' construction). This, heargues, implies that constructions do not have a single, fixed meaning, but thatthey instead are flexible schemata with abstract meaning potentials.

In 'Online changes in syntactic gestalts in spoken German' (pp. 127-155),Wolfgang Imo explores the question of how often garden-path sentences actuallyoccur in everyday interaction and whether they cause any problems ofunderstanding at all. He maintains that neither lexical nor syntactic ambiguityis a common phenomenon in modern German, potentially due to the disambiguatingrole of inflectional morphology. More importantly, real ambiguities do not seemto cause major communication problems since hearers always project multiplepotential continuations that even allow the resolution of apokoinu/pivotstructures in which two apparently independent constructions are merged into asingle utterance (cf. e.g. the blend in [da hat se GSAGT,]A [ja SCHÖNheit mussleiden]B [hat die KUH zu mir gsagt]C '[then she said] A [well beauty has tosuffer] B [this cow said to me]C' (p. 144, in which the sequences A-B and B-Care both well-formed quotative strings).

Structures that combine entrenched and open parts are the topic of SusanneGünthner's paper ('Between emergence and sedimentation'; pp. 156-185). Günthnerexamines three 'bi-clausal' constructions in German, pseudo-clefts, ''DieSache/das Ding ist …'' ('the thing/point is …') and extraposition with ''es''('it'), all of which start out with a relatively fixed, sedimented part, whichis followed by a fairly open schematic slot. These entrenched parts have a rangeof conventional functions, such as floor-keeping or the expression of speakerstance. On top of that, however, the first clause also has an important role inallowing the hearer to project the structure to come and helping the speaker inhis temporal, on-line discourse management.

Thiemo Breyer, Oliver Ehmer and Stefan Pfänder look at 'Improvisation,temporality and emergent constructions' (pp. 186-217) in situated interaction.They are mainly concerned with the creative adaptation, aka improvisation, ofpre-fabs in collaborative story-telling. As they point out, the specifictemporality of discourse opens up the possibility of joint language play. Thusdiscourse participants can pick up on previous utterances in the conversation orcommon cultural background and together improvise on these structures. Meansavailable to speakers for this include analogy and montage, i.e. the combinationof different elements into new and unexpected entities by e.g. blendingconstructional templates.

The next paper, by Peter Auer and Jan Lindström, 'Verb-first conditionals inGerman and Swedish: Convergence in writing, divergence in speaking' (pp.218-262) is the only one that contrasts constructions in spoken and writtenregisters. Drawing on comparative corpus data of German and Swedish, they findthat the conditional V1 construction is used frequently in written German, whileit surfaces only rarely in spoken German. In contrast to this, the constructionis widely used in both spoken and written registers in Swedish. As Auer andLindström argue, this difference in distribution can be explained by the factthat V1 structures in German are used for a greater number of alternativestructures and thus only allow for a weak projection of the conditional meaning.In Swedish, on the other hand, the construction is semantically much morefocussed on the conditional interpretation and thus allows for a more reliableprojection.

Dagmar Barth-Weingarten and Elisabeth Couper-Kuhlen take a close look at theinteraction of 'Action, prosody and emergent constructions' (pp. 263-292). Theycontend that it is not only the frequency and syntactic / semantic cohesion ofstrings that leads to their entrenchment. They suggest that togetherness inaction (as e.g. part of a turn-at-talk) and prosodic/phonetic form play acrucial role. Based on the findings on VP conjunction in a corpus of Americantelephone conversations, Barth-Weingarten and Couper-Kuhlen argue that thecloser the prosodic unity of VP conjunctions (with respect to e.g. intonationand timing), the greater the 'togetherness of action' (p. 264) and consequentlythe more likely the entrenchment of a conjoined VP string will be.

'On the emergence of adverbial connectives from Hebrew relative clauseconstructions' (pp. 293-331) by Yael Maschler is the last contribution to thevolume. Maschler notes that, according to prescriptive grammars, relativeclauses in Modern Hebrew must be introduced by the proclitic complementizer''she-'' and contain a resumptive element (pronoun, inflected preposition orverbal inflection, depending on the syntactic function that is relativized).Only in subject and object relative clauses is it considered acceptable to leaveresumptive pronouns unrealised. Yet, in his corpus of casual Hebrewconversation, he finds that it is adverbial complements that exhibit thegreatest proportion of such 'empty' pronouns. As a closer look at these tokensreveals, many contain a head noun that is fairly semantically empty (e.g. rega'minute'). According to Maschler, this, and the fact that ''she-'' is generallyused as a subordinator in Hebrew, has lead to a reanalysis of the antecedentnoun + relativizer ''she-'' sequence as a single temporal connective word thatintroduces a temporal adverbial clause.

EVALUATION

''Constructions: Emerging and Emergent'' is a coherent, well-edited volume thatshould be of great interest to linguists working with usage-based frameworks,Construction and Cognitive Grammarians as well as researchers working oninteraction and discourse analysis. All contributions show that a careful,qualitative analysis of authentic corpus data allows researchers to betterunderstand the formal and functional properties of constructions as well astheir use in discourse. This is particularly important since many ConstructionGrammarians, despite drawing on corpus data, still only analyse decontextualizedsentences. As the results from the various articles of ''Constructions: Emergingand Emergent'' indicate, however, such an approach might fail to uncover crucialfeatures of constructions such as e.g. their prosody or discourse managementfunctions.

At the same time, the purely qualitative approach taken by more than half of theauthors also has certain drawbacks. For example, Breyer, Ehmer and Pfänder'scontribution basically rests on the exemplary analysis of two conversationextracts, which obviously raises the question of whether or not one cangeneralize the findings from so small a database. Now, Breyer, Ehmer and Pfänderconsider their paper to be an explorative one, which might justify theirapproach. Many of the other papers, however, also face the same problem. All theexamples discussed do indeed seem to exhibit diverse functions such turn-takingand -yielding, eliciting hearer feedback, pre-disagreement and reproach or theexpression of speaker stance. Yet, one wonders how many of these functions arestatistically frequent enough to become sedimented. In order to answer thisquestion, I would suggest supporting the findings with quantitative corpusstudies. Note that this does not mean that I deny that single utterances canemerge by the creative assembly of constructions that have never been combinedbefore. As Hopper points out (and Günther convincingly shows in hercontribution), however, language use largely consists of combining existingpre-fabs (sedimented formulaic pieces with open slots and fixed expressions).Consequently, quantitative, statistical analyses would enable researchers tospecify how much of their data can be explained by, for example, discoursemanagement functions and how many tokens are in fact creative, one-off on-lineconstructs (this holds for the papers by Auer and Pfänder, Hopper, and PekarekDoehler) or how frequent garden-path ambiguities are per construction type (cf.the contribution by Imo).

On the positive side, four contributions to the volume already take such anapproach and also include quantitative data on their phenomenon underinvestigation (namely, Deppermann; Auer and Lindström; Barth-Weingarten andCouper-Kuhlen; Maschler). Of these, Auer and Lindström's paper is probably thebest example of how a careful combination of qualitative and quantitativeapproaches can yield insights into the use of a construction well beyond whatthe individual methods would have allowed. The only thing missing from theirpaper is a statistical analysis of their data. In fact, only one paper in thevolume, the one by Barth-Weingarten and Couper-Kuhlen, reports statisticalresults (and these, unfortunately, seem flawed: for their Table 1 (p. 280), theyfind two significant factor combinations, yet a chi-square test of theanalysable tokens in the table only gives an insignificant result of X-squared =4.39, df = 2, p-value > 0.10). Overall, this is somewhat disappointing in that,as argued above, the great number of potentially relevant variables clearlywarrant statistical investigation.

Despite these points of criticism, ''Constructions: Emerging and Emergent'' is animportant volume that combines discourse and interactional approaches with aConstruction Grammar perspective. It contains a great number of findings andraises many new questions that should spawn further research. Especiallylinguists working in the usage-based Construction Grammar framework should takegreat interest in the volume since it alerts them to the importance of formaland functional properties of constructions beyond the sentence-level.

REFERENCES

Aarts, Jan. 1991. Intuition-based and observation-based grammars. In KarinAijmer and Bengt Altenberg (eds.), English Corpus Linguistics. London/New York:Longman, 44-62.

Chomsky, Noam. 1965. Aspects of the theory of syntax. Cambridge, Mass: MIT Press.

Hopper, Paul. 1987. Emergent Grammar. Berkeley Linguistic Society 13. 139-157.

Hopper, Paul. 1998. Emergent Grammar. In Michael Tomasello, ed. The newpsychology of language. Mahwah, NJ. 155-175.

Hopper, Paul. 2004. The openness of grammatical constructions. ChicagoLinguistic Society 40. 153-175.

Sampson, Geoffrey. 2001. Empirical linguistics. London/New York: Continuum.

ABOUT THE REVIEWER

Thomas Hoffmann is Assistant Professor at the University of Osnabrück. Hismain research interests are usage-based Construction Grammar, synchronicsyntactic variation, and World Englishes. He has published articles ininternational journals such as Cognitive Linguistics, English World-Wideand Corpus Linguistics and Linguistic Theory. His monograph 'PrepositionPlacement in English' (2011) was published by Cambridge University Pressand the 'Oxford Handbook of Construction Grammar' (co-edited with GraemeTrousdale) will appear later this year. Currently, he is writing a textbookon 'Construction Grammar: The Structure of English' for the CambridgeTextbooks in Linguistics series.

Page Updated: 29-Mar-2012